A real-time architecture of multiple features extraction for vehicle verification
Citations Over Time
Abstract
In this paper, we propose a real-time architecture of multiple features extraction for vehicle verification. First, we set a range of YCbCr values to extract the pixels belonging to vehicle back lights. The density of light is computed by the number of the extracted pixels, and considered as the first feature. The second feature is the location of license plate. It is determined by Searching Area Decision, Local Edge Quantity, and Refinement. Then, the pixels belonging to back light and license plate are removed, and the averages of YCbCr values belonging to the remaining pixels are computed. It is considered as the third feature. Once the three features are computed, the relative error distance is applied to verify the vehicles in different frames. To achieve the request of real-time processing at 30fps, a four-staged pipeline architecture is also proposed. After synthesis, the total gate count is 390k and the operating frequency is 56MHz with TSMC 0.18μm process.
Related Papers
- → The effect of colour space on tracking robustness(2008)30 cited
- → Low complexity photo sensor dead pixel detection algorithm(2012)14 cited
- A Method Based on Rapidly Looking up Tables of Converting YCbCr to RGB in Video Decoder(2007)
- → Reversible Data Hiding Based on Multiple Strategies(2022)1 cited
- → A small-area and low-power circuit implementation of color space conversion from RGB to YCbCr(2016)